High-Magnetic-Field Research and Facilities: [Final Report] (1979)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

6 High-Magnetic-Field Facilities and Users INTRODUCTION To provide perspective on the current status and possible future develop- ments of high-magnetic-field facilities in the United States, it is useful to recall briefly some history. Because the saturation field of iron is 2 T, the first proposals for and achievement of fields above 3 T, as early as 1914, made use of iron-free high-power solenoids. Major advances occurred in the 1930's in the United States, with Giauque's work on kerosene-cooled solenoids at the University of California at Berkeley and Bitter's work on water-cooled solenoids at MIT. Bitter's basic design for high-field solenoid magnets has survived, and modi- fications of this design are used in most high-magnetic-field facilities today. Magnets of this and related designs are relatively inexpensive (^ $20,000 for a typical magnet); thus, the main capital cost of a facility is the power supply and cooling system. These features are responsible for the typical configuration of most of the high-field facilities in the world; namely, a number of solenoid magnet stations that can be connected on a time-shared basis to a single (or multiple) large power supply (supplies), providing a considerable flexibility in bore sizes, configurations, and the like. Recent rapid advances in high-field superconducting magnet technology have led to a substantial number of superconducting magnet systems pro- ducing fields greater than 12 T distributed throughout the world. These systems have provided certain additional capabilities not available with resis- tive magnets; however, very few of them are actually operated as true faci- lities. The highest steady fields (25-30 T) are being, or soon will be, produced by hybrid resistive-superconducting magnet systems, located necessarily at facilities with large power supplies for running the resistive magnets. For fields above 30 T, some type of transient operation, quasi-static or short- 72

High-Magnetic-Field Facilities and Users 73 pulse, is required at present. There are very few such systems in the world, particularly those producing fields of the order of 100 T or more, that are currently used for research. CONCLUSIONS The United States still leads the world in the number of large resistive magnet facilities, maximum field capability, number of magnet stations, and use of these facilities; but the development of facilities and their use in Europe have been increasing rapidly in the past few years. Recently, the increased demand for magnet time in both the United States and Europe has been for the highest available fields. A vigorous effort and substantial additional invest- ment in capital equipment will be required if the United States is to meet the needs in this field. Therefore, we conclude that additional capital equipment funds are required to accelerate the development of facility hybrid magnets and to extend the highest steady-state fields. Additional faculties should also be upgraded and expanded, as suggested in the section on Auxiliary Equip- ment and Facilities. The related questions of geographic distribution of present facilities and users and of establishment of small regional faculties have been considered. Currently, most large steady-field facilities are located in the northeast corner of the United States. Because of travel costs and experimental logistics, most users come also from the region where these facilities are found. A wider geographical distribution of users might be achieved through special travel funds allocated to facilities specifically for long-distance, shorter-term re- search projects. A complete resolution of these problems is outside the scope of this study, which is based on the view of only a small part of the con- cerned scientific community; however, the fact that the scientific oppor- tunities, discussed in Chapter 3, appear to be greatest in fields above 15 T, coupled with the capitalization and operational costs, argues against the estab- lishment of small regional facilities at present. Therefore, we conclude that small regional magnet facilities, with fields up to 15 T, should not be estab- lished at this time. No new individual high-field superconducting magnets should continue to receive support where the capital expense cannot be justi- fied scientifically; such magnets are not facilities and cannot be justified on that basis. Although several installations in the United States produce peak pulse fields in excess of 100 T, there are currently no facilities in the United States dedicated to the use of such fields for general research purposes. The tech- nology and expertise for producing short fields in excess of 300 T and quasi- static fields up to about 100T are available now. Transient fields are the

74 HIGH-MAGNETIC-FIELD RESEARCH AND FACILITIES cheapest and quickest way to achieve very high magnetic fields and the only way to achieve the highest fields of interest to researchers (~ 2500 T). In spite of the more limited range of experiments amenable to the transient-field approach and the difficulties in obtaining data, a number of workers have shown that practical experiments can be carried out. Therefore, we conclude that the establishment of a transient magnetic-field facility, or facilities, is required to meet the needs of research and technology for magnet fields in excess of 100 T. In determining the creation of the above-mentioned facilities, three fac- tors, in addition to capital cost, should receive consideration: 1. The quality and breadth of the scientific staff of the institution; 2. The geographic location, taking into account the current geographic imbalance in distribution of facilities and users; 3. The degree to which additional funding can be used to build on present technology. Further, any such facility should receive strong additional support for the development of sophisticated measurement techniques and additional equip- ment and facilities. The possibility of associating a small steady-field facility with the transient-field facility to concentrate effort and make the entire operation more cost effective should be considered. Therefore, we further conclude that facilities should be funded sufficiently that necessary support measurement equipment (e.g., optical spectrometers, ESR and NMR spec- trometers) be in place. CURRENT STATUS OF HIGH-MAGNETIC-FIELD FACILITIES Steady-State Field Facilities-Resistive Magnet Systems As we have mentioned, most of the high-magnetic-field facilities currently in operation in the world employ resistive magnets, and most of these are of the Bitter type and are water cooled. (The characteristics of the known resistive magnet facilities are summarized in Tables 4 and 5.) Other designs are em- ployed only in isolated instances; for example, the University of California at Berkeley in the United States and Oxford and Grenoble in Europe. Of the U.S. facilities described in Table 4, the Francis Bitter National Magnet Laboratory (NML) is the only one that serves the general scientific community. During the past two years, more than 120 different visitors made use of the high fields at NML for a variety of experiments. At the Naval Research Laboratory (NRL) facility, approximately 30 percent of the users

c cj 3 O J3 ^ "S £ «? D, X o" « -c o""o "o o ^_|^. o -g o" I ^? 2 ^53 - o. o O 6^ 6*i «^ u >, C C c c gt 0 <U u u o. ^^ i_ ^i^ o O 3 ^ u-i o ^^ S ^. O 00 S7'o | o| 0 ^ -S II o M 1 0 -H ~ \O (N y^j- oo 0. S o m C " 2 ^\t '" . _g h H" cT -H H S i/j H ro oo * oo 2 (V) ^ 0) OS "o s s — In e H C <U O <S S O ^Q 3 ^ .2 s 1 M C ^ cs J 3 00 ^~-' +-> a C* Z i ! IB IU & 0> ^_ C '"'^ 1 Q >-" S i-. r'C Q) fD O r ) ^ T3 "13 £« en £H ||| i i M o. ii ^ ij . ^ b, H . ^H A >*H GO JH 4-' O^ i S PQ £ • .a an g C « c 'S V Commissio Date O 00 oo vo o i 53 VO Tf 1 . i i-i en D Installation University of Pennsylvania University of California at Berkeley .Under construe Being phased o cExclusive of pe: Tt a i | H tu 2 75

SJ "^ 0 CO E 1 2" 0 £ e u E . o a E CL u E E I ^ E "u o ^. ,?" S o H v> o u 0 £ io_ si" S p> !£^ oo H" ro «1 SL !j» X H H H in' H O .^f H" H H CN (-H CD O u^ SO >^i in ^ 00 oo <N S 04 "B. o. is a C/3 0> |S S S u 22 * S^ '*"' iX .H m. S £ ^ S o "8 "~ So" u 2 S m' ft. * * S m o\ vo X 1 IN \ 3 u 4) X> 0) O. ^f ^2 -S >. i*5 ^5 A r { O T3 "o r 0 ^,, ^ tt'C «o "fl e u _>" S -rfw u .c u S w 0) M g U) eg S iS +." S 5 S H a £ £ S *-' Lebedev Institute of the USSR Academy of Sciences, Moscow High Magnetic Field Laboratory, Tokohu University, Sendai u High Magnetic Field Laboratory Grenoble (German participation) High Magnetic Field Laboratory Nijmegen §. Hochfeld Lab. der Technishen Universitat Braunschweig eg C.E.A., Saclay Clarendon Lab., Oxford International Laboratory for Magnetic Fields and Low Temperature, Wroclaw Kurchatov Atomic Energy Institute, Moscow 1 1 Installation h 2 "3 b o 4) a c« 1 !§> § y s o S? g o | M O 3 HH "o "§ _, " 00 — ^T P g e .•; S w £ 3 1 04 "o D .» 8 T3 8 1 1 "o h S « 'S S § "s i << § 1 "o o "3 8- B 3 H w u. u = o D ft, 0 * U. 04 CO 76

High-Magnetic-Field Facilities and Users 77 are external to the laboratory but use considerably less than 30 percent of the magnet time. The University of Pennsylvania and University of California at Berkeley facilities are used only by their respective university faculty and students, except for occasional outside scientific collaborators. Because the characteristics of the magnets and the necessary supporting equipment and staff are much the same, the general operation of most of these facilities is quite similar. Magnet "shifts" (typically 3% hours to 8 hours) are scheduled in advance; thus, users must set up experiments in advance and be prepared to run them during their shift. The estimated cost of a shift appears in Table 4. These numbers are rough approximations, but the estimates agree reasonably well (note that the University of Pennsylvania figure does not include personnel, maintenance, and supervisory costs, which receive separate support). The cost of electrical power constitutes a relatively small fraction of the total cost at present, with salaries and overhead of support personnel making up the largest fraction. A wide range of experimental studies are conducted at these facilities, including investigations of solids, liquids, gases, and plasmas, making use of optical spectroscopy, electrical and thermal transport, magnetic, ultrasonic, and MSssbauer resonance techniques, magnetostriction, and numerous other procedures and techniques. Most experimental investigations can be classified as solid-state physics or materials science. Table 6 gives a detailed breakdown of the types of experiments carried out during the past two years at NML and NRL. TABLE 6 Percentage of Magnet Use by Type of Experiment* Type of Experiment FBNML6 NRLC Superconductivity 32 22 Optics and spectroscopy 25 20 Magnetism 14 18 Transport (electronic and thermal) 11 20 Plasma physics 10 5 Miscellaneous (include resonance, Mossbauer, 8 IS liquid crystals, ultrasonics, biology) TOTAL 100 100 ^Percentage of generator hours. Past 6 quarters. cPast 2 years.

78 HIGH-MAGNETIC-FIELD RESEARCH AND FACILITIES Each of these installations provides additional facilities and equipment support to users. These generally include cryogenic fluids (at cost to the user), magnet monitor voltages, magnetic-field modulation, vacuum pumping, general-purpose electronic equipment (recorders, lock-ins, simple data- acquisition systems, etc.), and simple Dewars suitable for the different mag- net bores. Particularly at NML, and somewhat less at NRL, a substantial amount of specialized experimental equipment, such as spectrometers, lasers, and magnetometers, is also available, generally through arrangements with the scientific staff. This finding points to the essential role of the in-house scien- tific and engineering staff at the large facilities in developing specialized in- strumentation and state-of-the-art research capabilities for high magnetic fields. The use of these facilities varies considerably. At the NML, use has been increasing recently; at present, the demand for magnet time consistently ex- ceeds the available time, even with short shifts. The major increase has been in the highest-field (four-generator) magnets. At NRL, use has remained rela- tively constant for the past several years, averaging about 75 percent of the available experimental shifts. Use is less at the University of Pennsylvania facility, averaging about 40 percent. It is useful to compare the current situation in the United States with that in the rest of the world. Table 5 summarizes the characteristics of foreign magnet facilities. Recent developments in Europe are particularly interesting. Although the Clarendon Laboratory has had a resistive magnet facility for more than 30 years, the rest of the facilities are recent; for example, the joint French-German facility in Grenoble, the Braunschweig facility, and the Nijmegen facility have all been commissioned since 1971. (The Nijmegen hybrid magnet was built and tested at the NML.) In contrast, all U.S. facilities were commissioned prior to 1968. The joint French-German high-magnetic-field laboratory in Grenoble is the largest of the foreign facilities. We will describe it in some detail, for it provides a basis for comparison with the U.S. situation and illustrates the European commitment to high-field research. The Grenoble High Magnetic Field Laboratory was commissioned in 1972 and currently operates at a maximum power of 10 MW (two 5-MW generators), with six separate magnet stations. Two of these stations are capable of maximum fields of 20 T in a 5-cm bore, with two concentric Bitter magnets. The use of this facility has increased dramatically since its opening, from total magnet hours of 403 in 1972 to 4972 in 1977. Recent increases have been greatest in the 20-T magnets (each of which requires full power), placing even heavier demands on the facility. Forty to fifty outside visitors have used the magnets during the past year, in addition to the substantial German and French in-house scien- tific and technical staff. A wide variety of research programs has been carried

High-Magnetic-Field Facilities and Users 79 out at Grenoble; again the majority of these studies have been solid-state physics or materials science. Substantial additional facilities similar to those at the large U.S. installations but including high pressures and very low tem- peratures (dilution refrigerators under construction) are (or will be) provided to visitors. Steady-Field Facilities-Superconducting Magnet Systems Compared with that of resistive magnets, high-field superconducting magnet technology is relatively immature and has improved markedly in the past ten years. Many of the discussions concerning the relative merits of resistive versus superconducting magnets are based on older versions of the latter; however, magnets produced in the past several years offer greatly improved operating characteristics. Problems such as nonlinear behavior, degradation of field homogeneity with time, the need for retraining after thermal cycling, severe limitations on maximum sweep rates, and inability to modulate have been alleviated and, in some cases, completely eliminated. In view of this progress, we restrict our consideration to those facilities employing modern superconducting magnets and those capable of fields greater than 12 T (Nb3Sn or Nb3Sn plus V3Ga conductors). Magnetic fields only slightly lower than this can be achieved reliably and inexpensively with Nb-Ti multi- filament wire magnets, and facility-type operation is not an economic neces- sity for most workers. Table 7 shows the characteristics of the known modern superconducting magnet systems capable of fields of 14 T that are currently in operation throughout the world. [These magnets were all constructed by Intermagnetics General Corporation (IGC).] There are a considerably larger number of mag- nets with maximum fields in the range 12-14 T; but we will concentrate on the highest field solenoids (Table 7) to facilitate the direct comparison with resistive magnet facilities. The acquisition cost of these systems varied from less than $50,000 to well over $500,000 (the Japanese NRIM magnet system). A letter of inquiry was sent to all U.S. locations, and many of the foreign installations were also contacted requesting information about the operation of these magnets. Much of the discussion presented below is based on the responses to these inquiries. The magnets that seemed best (and most) utilized were those that were either (a) designed for a rather specific purpose, for example, split coil for optical experiments, and used fairly extensively by one or two groups for that type of experiment or (b) purchased as a general-purpose magnet that was part of a larger facility consisting of several lower-field superconducting mag- nets. Examples of the latter type are the Oxford magnet, which is a mobile

H 3 3 I i — — a 1 8 8 e s 5 o. Qi CO CO £ I o <a § E E p 2 o -> o o m ts 2 is >n "~: (N r- in' fi £, E P H S C H" H" S X in <*> t--_ o; c<^ \o a S * * m iX ^f' t' .a </) e 3 a o« u .S a- y u 3 3 "a S a 3 S X. O .2 53 . "2 o 2 Indiana University Sandia Laboratory University of Colo: Purdue University University of Calif' at Santa Barbara e b "a Location £ 2 2 o * « w £ "5 BO c u .<J o '3 H u 9 "g _« "3 e 5 .a" <u I OT O ^ _o z I "a o as E ?- 8 a . • c ^8 .S ^ II s 2 -a 5 8. s J3 E &" .S? £ u oo O o 0) 8 O If E o ^f ^f Tj" -*^ in .*6 .o \^ p o\ H" H H 8 H 2 -c 2'^ O I f> 22 Z 80

I § " E ^ 2 ^i *" 3 J= ^" O -a 00 2 u i .1« x a T3 - ° s Cu m O a> I S § £ JO 1 a, <D o, a) « b o H oq o arch Institute ate, Moscow ow Temperati X | £, o "o oscow 1 M O „ q) c "s OB « O o bJ rn Tokyo University Vac. Metallurgical Co" Japanese National Rei for Metal, Tsukuba Lebedev Physical Inst Electronics Institute, University of Wiirzbui Clarendon Laboratorj University of Sao Pau International Laborat Magnetic Fields and Kyushu University £ B Wroclaw o o pj 3 >> "o .| a c 60 o cd & H s5 \\ 13 « c "° "O i e S 0 B ^ o. "B "2 "o .3 g o i— i o b D ffi a. 81

82 HIGH-MAGNETIC-FIELD RESEARCH AND FACILITIES facility, literally "on wheels"; the Wurzburg magnet, where there are seven other lower-field magnets devoted to specific experiments; and the Sao Paulo magnet, which is also one of several superconducting magnets. The highest field superconducting magnet in the world (the Japanese NRIM magnet) has not been extensively used because of the general incon- venience (~ 2-h sweep to full field) and high cost of operation. During the past year only about ten shifts were utilized. The only U.S. magnet system that seems to have been purchased with the intent of making it a true facility magnet is at NRL. This system employed a "foot-type" room-temperature access Dewar; it is intended to keep this system cold for several months at a time, initially with batch filling but soon with a closed-cycle refrigerator to reduce operating costs. All other respon- dents to the inquiry from the United States indicated that only batch filling was being used. The NRL magnet system is not yet fully operational; there- fore, this type of facility operation cannot yet be assessed. From the survey, we conclude that: 1. Few of these magnets are actually operated as facilities; typically only one or two groups use the magnet, and use is generally rather low. 2. In several cases, operating costs (generally cryogenic fluids) have dis- suaded extensive use of the systems. Evaporation rate and cryogenic design are of concern to many. 3. Geographic location and resulting travel costs and specialized re- quirements (stability, high homogeneity, sensitive experimental equipment, etc.) were the principal reasons for purchasing the superconducting magnet rather than using a central facility. 4. Single superconducting magnets cannot do all things for all people. A single superconducting magnet is not a facility and is not cost effective in this context. However, individual high-field superconducting magnets can be ex- ceedingly effective and useful when dedicated to specific experiments if the capital and operating costs can be justified. Steady-State Facilities-Comparison of Resistive and Superconducting Magnet Facilities The advantages and disadvantages of resistive and superconducting magnets are discussed in Chapter 5 of this report. Here we provide an economic comparison of the two types of facilities (a hypothetical facility in the case of the superconducting magnets, for a true facility does not exist at present), considering first the acquisition cost, then the operating cost of each type irrespective of their inherent advantages and disadvantages. It is not valid simply to compare the acquisition and operating cost of a single high-field superconducting magnet with acquisition and operating costs

High-Magnetic-Field Facilities and Users 83 of a large resistive magnet facility, or for that matter, to compare costs of running single comparable resistive (electricity costs) and superconducting (liquid helium and nitrogen costs) magnets. A more meaningful comparison would be total costs of similar facilities that could provide a comparable range of fields, bore sizes, and the like. For the resistive magnets, we choose a modest facility similar to that at NRL, that is, a 6-MW rectifier power supply, six stations, and a mix of three bore sizes (1$, 2\, and 4^ in.) with a maximum field of about 1ST. There are three full-time technicians and one full-time professional associated with the facility. To provide a comparable (hypothetical) superconducting magnet facility, we take three separate magnet systems, one l|-in.-bore 17-T magnet and two larger-bore 15-T magnets, perhaps with special configurations for optical access, high homogeneity, room-temperature access, and the like. Addi- tionally, we consider 1.5 full-time technicians and a half-time professional to be associated with cryogenic aspects, measurement technique development, scheduling, maintenance, and administration of the facility. The acquisition cost of the resistive magnet facility is estimated as follows: power supply, $600,000; cooling system, $500,000; magnet design and construction (six magnets plus three spares), $300,000; miscellaneous equipment, $200,000; total, $1.6 million. For the "comparable" superconducting magnet facility, we estimate the following costs: magnets, one 16.5 T at $350,000 and two 15 T at $175,000, that is, $350,000; miscellaneous equipment (refrigerators, measurement equipment, etc.), $150,000; total, $850,000. Thus the capital costs of the resistive magnet facility are about a factor of 2 greater than for the super- conducting facility. Operating costs are estimated for the resistive magnet facility from the actual total costs of the NRL facility, including salaries and overhead, main- tenance, and the like, for 75 percent use of the total number of available 5-h experimental periods of $ 1500/period. The costs of the superconducting magnet facility are estimated as follows: 1. The total salary and overhead for two man-years, $130,000. 2. Batch filling, two experimental periods per day per magnet, typically two magnets running simultaneously, 12 cool-downs per year per magnet, helium and nitrogen costs of $3.25 per liter of liquid helium, and an average evaporation rate of 1.5 liters per hour; total cryogenic fluids cost, $91,000 per year. Thus, with 780 experimental periods (of 5 h each) used, the cost is about $300 per period, significantly less expensive than the resistive facility.

84 HIGH-MAGNETIC-FIELD RESEARCH AND FACILITIES In projecting this economic comparison into the future, one must take into account that both liquid helium costs and the cost of electrical power are expected to increase substantially during the next several years. Steady-State Field Facilities-Hybrid Systems The design of and rationale for hybrid magnets in extending the capabilities of resistive magnet facilities are discussed in Chapter 5. There are at present operating hybrid magnets at the Kurchatov Institute (25 T), Nijmegen (25 T) (designed, built, and first operated at the NML), and the Clarendon Laboratory (16 T). Hybrids are under construction at the NML (30 T) (30 T has already been achieved with 10 MW and the Nijmegen magnet superconducting solenoid) and NRL (24 T), and one is planned at Grenoble (30 T). The estimated cost of the 30-T magnet at NML is $500,000; the projected cost for the planned Grenoble magnet is 10 million French francs ($2.25 million at the current exchange rate). The large disparity in cost appears to result from the inclusion in the Grenoble estimated magnet costs, of person- nel costs for the extensive design and development for both the super- conducting outer solenoid and the polyhelix design resistive inner coil over about a 5-year period. At the NML, on the other hand, the development costs for both the resistive and superconducting solenoids have been largely written off in producing the Nijmegen hybrid, thus the costs represent only actual construction costs. In addition, the Grenoble system is to have a 5-cm bore, necessitating a 13-T, 50-cm-bore superconducting magnet; while the NML hybrid has a 3.2-cm bore and requires a correspondingly smaller super- conducting magnet section (7.5 T, 40-cm bore). Transient-Field Installations-Introduction Well-developed, conventional high-field instrumentation can be used for field measurement, temperature measurement, and various types of experimental data acquisition in steady magnetic fields. In transient fields, on the other hand, particularly for the shorter time scales, measurement techniques are not so well developed; signal to noise is a much greater problem, and equilibrium conditions are not generally established. In addition, steady fields are easily used by inexperienced researchers, whereas the problems just cited for transi- ent fields require a much more detailed knowledge of the technology and measurement techniques. Transient Field Installations Quasi-static Systems Certain types of magnets are designed to be capable of functioning con- tinuously at high power but receive their power from a limited energy source.

High-Magnetic-Field Facilities and Users 85 One magnet system of this type is installed at the Australian National Uni- versity in Canberra; it is powered by a homopolar generator that stores 580 MJ. This system has provided fields of almost 30 T in the 5-cm bore of a polyhelix magnet for several seconds but is inactive at present. This field is not pulsed in the usual sense, for the heat capacity of the magnet cannot absorb the Joule heat, and cooling similar to that of steady-field resistive magnets is required. Advances have been made in recent years in other types of quasi-static field generation. Fields up to about 40 T have been generated and maintained for large fractions of a second. At the University of Amsterdam, a liquid- neon-cooled solenoid produces constant 40-T fields (within 0.1 percent) for 80 msec, with a 6-MW rectifier supply. The coil starts at 30 K and heats to about 150 K after each pulse. Approximately 2 h are required for the magnet to cool again, leading to a duty cycle of about 1CT5. In Toulouse, France, a different approach has been used to generate quasi- static fields. This system is described in Chapter 5. It has been used success- fully as a facility, and a number of different visitors have used the field for experiments in solid-state physics, usually on a collaborative basis. Quasi-static fields have been generated in the United States only at NML. Transient-Field Installations-Short-Pulse Systems Nondestructive pulsed fields obtained by capacitor discharge have long been used to produce fields between 30 T and 100 T for periods of milliseconds to microseconds in small volumes. These are generally rather inexpensive sys- tems that require modest capacitor banks; therefore, they have not come into widespread use as facilities. The range of experiments in this region is even more restricted. Relaxation times must be short, and signal to noise must be inherently good, because integration times are necessarily very short. For useful fields above 300 T, some type of flux compression scheme is necessary, either electromagnetic or explosive. Destruction of the coil and/or sample is a serious drawback to these approaches. However, in the past there has been significant effort devoted to production of ultrahigh magnetic fields by these techniques, and recently there has been renewed interest because of possible applications in controlled fusion research and compact, electrical power-source development. A "fallout" from these efforts is the development of techniques for field measurement, as well as signal-processing techniques; the latter have also benefited from short-pulse work in other fields, for ex- ample, short-pulse laser technology. Ultrahigh magnetic fields are produced at a number of installations in the world. In the United States these efforts are concentrated around magnetic- field confinement in fusion research, and fields are not usually produced for general research purposes (see Table 8).

u 0 "J 0 0 2 1 u ration iif o o u o 3 i« CO <U O *> "* w 'J IS s u C "eel o 1 1 V i o O Q * S. .* f> p- 1> S-. Si e "rt " £ T? 1 PP0 i S ° a" § _^ 8 "3 V O O O iy^ 17-1 i/~i — " P-J — o o > - 2 7 ii. "". H P H P" o P M cd Q 0 1 1 1 O in O H 5" ^ o ^ H H o u O o o r- o o UH ^ ^- W) A — V o o 00 O o CU CM — — o <n ^> ^^ « ^^ 2 C O 'vS « ct) S a en DO JD (U O o C O s 1 ^ « .x JO ,'^ on y, u Q CJ c u D. G, _^ « 0 .a « £ 'c Q. '3 <-. ,— . E -D c o ^ ±- £ c c "c U a o 'S fll cd ^ i * S O C cd Cd "-, IM e -° S t^ o. 0 ^^ ^ -G 0 o S c "5 S > ^ S 0 0 O <Q 8g »^ o O r- « o i o a 1 "§ 2 g x '3 'o cd ^ 1 § u s 0 O 0 u £ ca 'Tf £" o "" § S" i I/) IN £ 2 x ^ ^ & ^ o e >- c« o" fN OJ VI ^ O o .° E * S I " g H .° ^ -° 4 ji 2 >, EC ? C &0 « ^" C ",-? 1 0 M V) X ra C, x x x o x 0 0 O -^ «> O O rj i E o 'C, 6 5 E S rs E E >-. OJ X « c ° T 7 U u .C LL o O o <J £ "" l -~. **. M rg / / fS V) fN O (N E/> Q. 3 ° C e cd e e i 5? u U S i £ V 0) a C .6 a> a> ,i u B*l S -§ 0) > (U i- 2 0) i o tj o w .O E u T3 > u "*2 "O u C > u CU n) « u U o e e < < £ < < D < i flS n C v e _] -2 "^ E G « -g 1/5 -3 1 g !« !H ^ 5 rldwide Trans Installation Lawrence Li M Amsterdam Australian N Universitj Institute for Physics, I of Tokyo O 3 PBNML Toulouse "o M o «> e o Jj « 00 1 a >, -3 3 j jH "o C S e u to e " JS g ,j 0 'e "o =1 2 H g 3 £ < »i — 86

High-Magnetic-Field Facilities and Users 87 Destructive ultrahigh fields up to 280 T have been produced at the Insti- tute for Solid State Physics, University of Tokyo, and used for research in semiconductors and magnetic materials, as well as for development of meas- urement techniques for very high (and destructive) magnetic fields. Operations at facilities at Frascati (Italy), Grenoble (France), Tohoku Uni- versity (Japan), and Illinois Institute of Technology have been suspended. FACILITY PLANS AND POSSIBILITIES Current Plans-Steady-State Fields As far as we can determine, no plans exist at present for extending the maximum fields produced by resistive magnets beyond those now obtainable at NML. However, construction of hybrid systems is planned or already in progress at all of the large resistive magnet facilities in the world that do not have such systems. In one case, NML, operation at 30 T has already been demonstrated. Thus steady fields of 30 T should soon be available for users of NML in the United States and in several years at the High Magnetic Field Laboratory in Grenoble for European users. A 20-T superconducting magnet is planned jointly by IGC and the Univer- sity of Tokyo, but several years will be required to develop, construct, and test it before it can become operational. It will also probably be quite costly (see Chapter 5). Current Plans—Transient Fields A quasi-static field facility that will be capable of over 30 T for times up to 1 sec is being developed at NML. This system utilizes the present dc genera- tors (2) with a bank of 20-kA SCR switches. A prototype has been demon- strated. Currently, we know of no plans to develop short-pulsed field facilities for general research purposes in the United States. hi Japan at the Institute for Solid State Physics, University of Tokyo, there are plans to develop a quasi-static system (10 to 100 msec) capable of up to 70 T for metals studies. Additional plans include the development of a short-pulse system using a 5-MJ condenser bank in an effort to produce a maximum field of approximately 1000 T in a 0.8-mm-diameter space. In Louvain, Belgium, the establishment of a short-pulse facility for fields in excess of 100 T for general research purposes has been planned for several years. Implementation has been hampered by lack of funds.

88 HIGH-MAGNETIC-FIELD RESEARCH AND FACILITIES Future Possibilities-Steady-State Fields It is clear (from Chapter 5) that the most efficient and cost-effective means of providing steady fields higher than those currently available on a facility basis will be through the use of hybrid magnets. These facilities can be developed most cost effectively at locations where resistive magnets and their power supplies already exist. In the near term, that is, from 2 to 5 years, 30 T will be available at the NML and, with a modest additional investment in a higher- field superconducting booster coil, this capability could be extended to 35 T with the present power supplies. In the longer term, from 10 to 20 years, a substantially larger investment could allow facility operation of a hybrid magnet up to 50 T (see Chapter 5); perhaps, with further technological devel- opment, the 75 T envisioned in Chapter 3 could be attained. Future Possibilities-Transient Fields-Quasi-static Systems The cheapest and quickest way to achieve fields in the range 50-100 T that could be provided to users on a facility basis is through the use of quasi-static systems. With present power supplies for fusion experiments, peak fields in the 50-75 T region, with durations of several tenths of a second, can be generated. One such power supply is being installed at the NML. Application of the principles determined from the development of the current quasi-static system should permit extension of the maximum fields to approximately 70 T with this generator. Although there are a number of drawbacks of transient fields (even of this relatively long duration), they appear to be the most cost-effective means of providing high fields in this range for the majority of users, for the cost of the magnets is relatively low ($100,000 for a 50-T magnet) and somewhat smaller power supplies are required than for steady fields. Future Possibilities-Transient Fields-Short-Pulse Systems Future possibilities for extending the maximum field, duration, and volume of the various types of short-pulse magnets were discussed in detail in Chapter 5. The magnets are inexpensive, but they must be replaced fre- quently (after each shot for destructive operation). High-voltage, fast- capacitor banks can be constructed with an investment of about $1/Joule, or one of the large capacitor banks at, for example, Los Alamos or Lawrence Livermore Laboratory could be used. With the latter approach, further re- search and development could be undertaken with a relatively small addi- tional investment. On the other hand, facility operation of magnets of this type presents a challenge. A somewhat restricted range of experiments can be performed; optical measurements or particle beam probes appear to be best

High-Magnetic-Field Facilities and Users 89 suited for this type of field, for the probe (source) and spectrometer/detector can be completely external to the magnet. Thus, this type of facility should provide a wide range of intense EM radiation sources (particularly lasers) to span the spectrum between the microwave and ultraviolet, as well as fast detectors and spectrometers (rotating-mirror spectrometer cameras with image intensifiers). Other techniques that have proved successful are dynami- cal susceptibility and magnetoresistance; equipment of this type, as well as the sophisticated fast electronics, timing circuitry, and reliable field measure- ment techniques, would also be necessary. Therefore, a substantial investment ($1 million to $2 million) would be required for these necessary auxiliary facilities and techniques. FACILITY UTILIZATION AND USER CONSIDERATIONS Utilization Many factors contribute to effectiveness and use of a high-magnetic-field facility. To a large degree this problem is circular, and this fact should be recognized. The utilization of facilities depends on physical location, maxi- mum fields and configurations, additional facilities provided, dissemination of information about the facility, and the like. Yet, development of new facili- ties and increased capability for present ones depend on demand from the scientific community and/or recognized payoff in new scientific and tech- nological opportunities. Any sort of realistic projection of use is difficult because of the factors that we have just mentioned and could, in fact, become self-fulfilling. With this in mind, we do not attempt to offer a detailed projection of future use of high-magnetic-field facilities but only reiterate that use of the NML has shown dramatic increases recently (at present 100 percent) and that the large European facilities, such as Grenoble, also show a steady increase in demand for magnet time over the past few years. Full utilization of the NML facility, as well as of the Grenoble facility, results in part from increased demand for the highest-field magnets that use the entire facility power, thus precluding simultaneous runs. Most of the increased scientific interest is in the highest available magnetic fields. The discussion that follows is based largely on information about and experience with the large resistive magnet facilities (most of which relates to NML), with some input from superconducting magnet "facilities"; at present no similar data base for transient-field installations exists. However, some of the discussion is sufficiently general that it could apply to any type of high- field facility.

90 HIGH-MAGNETIC-FIELD RESEARCH AND FACILITIES Geographic Location and Travel Physical location and resulting necessary travel for many investigators are factors affecting use of high-magnetic-field facilities. The two major resistive magnet facilities (and three of the four such facilities) are located in the northeast part of the United States. It is useful to examine the geographic distribution of the outside visitors to the NML (including long-term visitors and some from abroad, though a small number of these visitors do not use the resistive magnets). Figure 7 shows graphically for a two-year period the per- centage of outside visitors versus the distance of their home institutions from the NML. More than 37 percent of the visitors are located within 100 miles of the NML (mostly in the Boston area). It appears that travel is a limiting factor not only because of its cost but also because of the extra time and logistics required for an experiment. Travel cost and living expenses for a group in- cluding a senior investigator and several graduate students and coming from an institution located more than 100-200 miles away can be a significant limitation. Equipment of groups coming from greater distances generally is not transported back and forth but left at the installation on a semi- permanent basis. For most standard research grants, the cost of one or two such trips can represent a fair fraction of the total travel budget if the high- VISITOR GEOGRAPHIC DISTRIBUTION NM L January I976 to Apr!I I978 200 400 GOO 800 1000 1500 2500 00 DISTANCE (Miles) FIGURE 7 Visitor geographic distribution, January 1976 to April 1978.

High-Magnetic-Field Facilities and Users 91 magnetic-field experiments are only a part of a larger research program that is conducted predominantly at the investigator's home institution. If the experi- ments are planned far in advance, specific funding can be requested, but for shorter-term efforts, costs can be a real problem. Many investigators cannot afford to travel great distances, and, even if travel funds are available, the logistical problems are a strong deterrent. A few additional magnet facilities could not substantially alleviate the problems except for a small number of potential users who would be located from 100 to 200 miles from such a facility. Application and Screening Procedures Scientists who might be interested should be informed of the existence of high-magnetic-field facilities and of the details of fields provided and addi- tional available facilities. Potential users can learn about high-field facilities in several ways: from journals that publish work conducted at an institution or that feature news stories (for example, Physics Today); by informal interaction with those who have visited the facility; through personal contacts with members of the staff of the facility; and through descriptive brochures (both the NML and the NRL have recently sent a large number of brochures to members of the scientific community). Screening of potential users of the NML is carried out in several steps. After an initial verbal screening comes a mandatory written application; the written application form is sent to about 75-80 percent of those who apply. The large majority of written applications are approved, with the result that about half of all those who have applied (verbally or in writing) to NML eventually carry out experiments with the high-field magnets. No attempt is made to assign priorities to the applications according to scientific merit. The feeling is that even if the scientific value of an experiment is doubtful, or the competence of the experimenter, or the need for the high fields, approval should still be granted as long as only a small amount of magnet time is involved. When a large amount of time has been requested, the application is approved for a modest number of shifts, with the work then being re- evaluated on the basis of the initial results. The feasibility of an experiment frequently is not in question, for many potential users have lower-field mag- nets at their own institutions and have already conducted the proposed ex- periments in those lower fields. A Users Committee has been established to resolve conflicts, oversee allocation of facility time, and consider facility improvements for users. At the NRL, as the number of outside users is relatively small and magnet time is not fully used, no formal application and screening procedure has

92 HIGH-MAGNETIC-FIELD RESEARCH AND FACILITIES been established. Requests are generally made through staff members, and no additional screening is done. Scheduling Scheduling magnet time at a large facility can present problems, especially when time is oversubscribed as it is currently at the NML. Short shifts and careful scheduling (every two weeks) have provided a partial solution; how- ever, a substantial queue still exists. In addition, although the time pressure and need for detailed planning might contribute to more efficient experi- ments in a given time period, there is less time to think and to assess inter- mediate results; this situation can result in some wasted effort and inefficient use of magnet time. Further, large blocks of time are quite difficult to obtain. Because part of the scheduling problem has been caused by increased demand for the highest fields, considerable relief would be provided by installation of hybrid magnets as part of the facility. For example, the same 10-MW power supply could be used to run one 30-35 T magnet, two (or perhaps three) 25-T magnets, or three (or four) 20-T magnets, provided the suitable super- conducting outer sections were available. The NRL uses a simple weekly scheduling procedure; requests for magnet time the following week are made at a weekly meeting. As the available time has not been fully utilized, there are generally few scheduling problems, and with sufficient notice, large contiguous blocks of time (up to a full week's run) can usually be obtained. Again, no assignment of scientific priorities to experiments is carried out. The possibility of establishing scientific priorities for the use of magnet time to alleviate problems in the demand for available magnet time has been considered; the general feeling is that because of the broad range of research projects and the usually short time scale of the experiments, such a procedure would be difficult to implement and, except in rare cases, would cause more problems than it solved. Auxiliary Equipment and Facilities All the large resistive magnet facilities provide additional technical support, instrumentation, experimental equipment, and, in some cases, additional ex- tensive facilities. We are concerned here with possible improvements in such facilities and the need for specific additional equipment and facilities. A problem often encountered in the use of high-field facilities is the neces- sity for repeatedly assembling and disassembling highly complex and sensitive experimental instrumentation at the magnet stations. When the facilities are not fully utilized, apparatus can remain at a magnet station for relatively long periods with only minimal difficulties; however, when there is a queue of

High-Magnetic-Field Facilities and Users 93 users, the assembling and dismantling of equipment can lead to substantial inconvenience for a large number of users. The addition of more "popular" magnets and magnet stations could partially alleviate this problem. In addition to the routine equipment and consumable materials (cryogenic fluids, for example) that are generally provided by high-field facilities, the availability of certain specialized equipment that is either too sensitive or too large to be transported easily would make high-field facilities more generally useful and might attract scientists who, for reasons of logistics and equipment problems, have had to limit their use to the lower fields provided by super- conducting magnets. Some examples of such specialized equipment are general support facilities for chemistry, ultrahigh vacuum chambers and heat pumps for chemistry experiments, a complete range (submillimeter to ultra- violet) of lasers (including dye lasers), spectrometers and detectors (with appropriate optical components and hardware), pulsed lasers with associated detectors and signal-processing electronics (particularly useful for short-pulse magnet systems), signal averagers, and digital acquisition and data-processing equipment. On a somewhat larger scale, facilities such as ESR and NMR spectrometers and very-low-temperature cryogenic facilities (dilution refrigerators) would be very useful in relation to some of the scientific opportunities discussed in Chapter 3. Much of this equipment is commercially available, but some of the very specialized equipment, such as NMR facilities and dilution refrigerators, will only be available to and truly usable by visitors through the efforts of an active, innovative scientific staff associated with the facility.